Life Beyond the Briefs

Using AI to Make a Solo Practice Scalable | Carolyn Elefant

• Brian Glass

Share episode

What happens when one of the original voices in solo law practice meets the cutting edge of AI?

At 8AM’s Kaleidoscope event, Brian sat down with Carolyn Elefant - legendary founder of MyShingle, author of Solo by Choice, and environmental justice attorney fighting Big Energy with bold strategy and even bolder tech.

In this conversation, Carolyn shares how she:

  • Turned a dot-com-era blog into a pipeline for high-value clients
  • Uses AI to decode 200-page rulings in minutes (and train new lawyers fast)
  • Created a playbook so effective, government agencies distributed it for her
  • Is building an AI-first law firm offshoot with scalability in mind

This isn’t your average solo practice story. It’s a deep dive into how to work smarter, build authority, and design a law firm that actually fits your life — not the other way around.

Whether you’re solo-curious, tech-savvy, or just sick of the billable grind, this episode will make you rethink what’s possible.

Hit play and get inspired by Carolyn’s unapologetic, future-facing approach to law.

Connect with Carolyn:
www.myshingle.com
LinkedIn: https://www.linkedin.com/in/carolynelefant/

____________________________________
Brian Glass is a nationally recognized personal injury lawyer in Fairfax, Virginia. He is passionate about living a life of his own design and looking for answers to solutions outside of the legal field. This podcast is his effort to share that passion with others.

Want to connect with Brian?

Follow Brian on Instagram: @thebrianglass
Connect on LinkedIn

SPEAKER_00:

Hey friends, welcome back to a special kaleidoscope edition of Life Beyond the Briefs. I recorded this next conversation live at the 8 a.m. Kaleidoscope event, and it's one I've been looking forward to sharing with you for a while. Today's guest is Carolyn Elephant. If you've been around the legal innovation space for any length of time, you already know that name. Carolyn is one of the original voices advocating for solo and small firm lawyers to build something on their own terms. She launched MyShingle.com back in the dot-com era, yes, before Legal Influencer was even a thing. And she's been ahead of the curve ever since. In this episode, we dive into how she built a niche practice fighting for renewable energy, landowners, and communities, and how she's now using AI tools to scale that work more efficiently than ever. We're talking blogging before blogging was cool, getting referrals from pipeline lawyers who don't want to be named, and why she she's building an AI forward spin-off firm in 2025. If you're wondering how to stay relevant in a rapidly changing legal world without burning out or selling out, Carolyn drops a masterclass in this conversation. She's sharp, honest, and probably using AI tools in smarter ways than most firms even know is possible. Let's get into it.

SPEAKER_01:

Hey guys, welcome back to the show. Today I have one of the OGs in the law firm business space, Carolyn Elifant. Back before there was a legal build your law firm conference every weekend, and sometimes two on a weekend, you were running the MyShingle blog, and you were one of the original voices in the space. So I'm curious, what gave you the idea to start blogging on the internet about building a small law firm?

SPEAKER_03:

So I started so long ago that it was during the dot-com era when companies were starting paper, uh starting companies using napkins, you know, or starting companies like napkins.com, and I thought, wouldn't it be great if there was a site that was dedicated to solo and small law firms? The problem was is it was such a long time ago that it was really costly to build a website back then. And so I waited for like a year on that idea, and then blogging became something that was popular, and so I started as a blog, but the problem with blogs was that they weren't revenue generating back then, so it became a free site for a long time, a free site that wasn't read very frequently either.

SPEAKER_01:

Aaron Ross Powell Oh, I don't know about that. I mean your your name came up over and over and over again in that um space back in 2002, early 2000s. And then of course you wrote the book solo by choice. It was an ABA published book.

SPEAKER_03:

It's not the ABA, but uh it was the ABA I think did market it or send it out at one point in time.

SPEAKER_01:

Aaron Powell And all the while running your own solo law firm.

SPEAKER_03:

Yeah, yeah. I did those all the time.

SPEAKER_01:

Fighting climate change.

SPEAKER_03:

Right, right.

SPEAKER_01:

Aaron Ross Powell Tell me about that practice because that's you know, I come to a lot of these conferences, and most people are in personal injury, trust in estates, criminal defense, immigration, a lot of immigration lawyers who are kaleidoscope. I've never met somebody running a small law firm fighting climate change. So what do you do?

SPEAKER_03:

Aaron Powell So I work with I do it in two ways. I work with renewable energy companies and help them sell their power into the grid so that it's more green friendly. And so they're basically fighting against larger incumbent utilities, kind of like solos and smalls are fighting for a space in the uh larger uh big law incumbent playing field. And then I also work with landowners and communities that are impacted by gas pipelines and bioelectric transmission lines, and they either fight those projects or try to get more revenue out of them or try to make them more accommodating to those communities.

SPEAKER_01:

Aaron Powell And one of the things you're presenting on uh at this conference is using AI to scale your marketing, scale your law firm. How do you find clients as somebody in in the space that you're in? Or how do clients find you?

SPEAKER_03:

Trevor Burrus, Jr.: So it's very interesting. Many years ago I came up with what I thought was sort of the uh a silver bullet for doing that. I put together this whole comprehensive guide for communities on how to fight pipelines or what to do in order to l level the playing field. And I got very lucky because all these landowners were calling the agency that regulates these projects and asking for help. And the agency actually circulated this PDF file that I had put together. And soon all these landowners started sharing it with each other. And so basically somebody else was doing the marketing for me. So that's where I get one set segment of the clients. And then the other thing is is opposing parties will often represent, will often refer people to me. But the problem is that they do it in secret. They don't want to be known as the company that referred somebody to Carolyn Eliphant. So I can't really thank them for the referrals or really keep track of it. But I do know I've gotten many anonymous referrals over the years.

SPEAKER_01:

Aaron Powell I mean it's almost like the imminent domain practice space where there's, you know, there's a state actor coming in and buying up plots of land. And hey, before you consider this offer, maybe you should go and talk to this person. Is that how you're getting referrals? There's anonymous secret referrals from developers or from pipeline builders?

SPEAKER_03:

Aaron Powell Yeah, that's that's exactly how from the council who represent them. And there are also like there is a very large eminent domain component in this because these projects have the power of eminent domain. So it's yeah, that's really an apt analogy.

SPEAKER_01:

Aaron Powell Isn't intentional that you got your white paper in front of that agency that now distributes it to everybody that asks?

SPEAKER_03:

Aaron Powell No, it was completely accidental. I was shocked that they were doing that for me.

SPEAKER_01:

Aaron Powell This is the power of creating your own authoritative branding content and getting it out there into the world. Sometimes somebody who you don't even know about picks it up and starts distributing it to your client base.

SPEAKER_03:

Trevor Burrus, Jr. Yeah. And I think what one thing people don't understand is that if you want that to happen, you really have to produce content that has value. It has to have more value or as much value as something somebody would pay for. I mean, if this had just been like a list of bullet points or like Captain Obvious points like don't talk to the company without an attorney, you know, something like that that was stupid, I don't think it would have gotten the traction that it did. But it had, you know, citations to cases. It had sort of one section that was very user-friendly and readable, but then the back I had all the source material with cases and regulations. Because I knew also that many landowners like to read that stuff, but that gave it credibility. Trevor Burrus, Jr.

SPEAKER_01:

And you know, there's many people out there that would tell you you're giving away too much information.

SPEAKER_03:

Yes.

SPEAKER_01:

And you shouldn't. But it's like I can't i in a in a personal injury case, which is what I do, I could give you a bunch of information you could probably negotiate with an insurance company. I can't imagine picking up that paper, even with all the citations, and being able to defend myself against somebody who's building a pipeline through the backyard. So I imagine that that that authority building for you then converted clients into going, you know, this is really hard for America to do it for me.

SPEAKER_03:

Aaron Powell Yeah, exactly. It happened that way. And sometimes clients will there's parts of the process they they can still handle on their own, but they're handling it with the guidance that I provide in the book. So when things get to the point where they reach the point where you really do need to ret retain an attorney, at least they haven't botched the case beyond recognition. Trevor Burrus, Jr.

SPEAKER_01:

That's that's the goal with providing information is enough information that you can handle maybe a smaller case on your own, but not enough that you can totally screw up the case for one week. How big is the firm now?

SPEAKER_03:

Aaron Powell It's gone up and down in size. There was a point where I had two associates. Now it's just me and my virtual assistant who I've had for about 17 years. And then I just bring people in on a contract basis. So for example, right now I'm handling a large transmission project in Maryland, and I've brought in one attorney on a contract basis, and they bring in another. And so I just sort of staff it very dynamically. I have a good network of people who I can bring in and bring on and off cases. Some are retired attorneys who've worked in the energy space, some are new attorneys who are trying to get into the energy space. So I kind of have a good wealth of people I can draw on. And now because of AI, they don't have to be as experienced as they once were, because that helps to level the skills gap. Aaron Ross Powell, Jr.

SPEAKER_01:

All right. Tell me about that. How are you using AI to speed up the rate with which your lawyers are learning?

SPEAKER_03:

Aaron Ross Powell So one of the things it can do in energy, there's just so much jargon, so many acronyms and so much institutional knowledge that you need to have. So I have lots of information on how energy practice works, and AI can help kind of guide people through that by summarizing it and giving them the kind of background that they need to be able to get up to speed really quickly. And then it also helps cut through. I mean, some of these orders in the cases that have um precedential value, they could literally be 200 pages long. So instead of having, you know, paying somebody eight hours to read through a 200-page order, they can use sort of an annotation that AI puts together and get through it in half the time. So those are just some examples.

SPEAKER_01:

Is this are you using where you get ChatGPT or is it something you've custom built, or is it a vendor Aaron Ross Powell?

SPEAKER_03:

I use all the all the different publicly facing tools. So ChatGPT, Claude, Gemini, it really just depends on the on the project, depends on my mood. I've used some of the vendor research tools, but they haven't really given me much more of an advantage in terms of getting people up to speed or summarizing cases.

SPEAKER_01:

Aaron Powell And the big boogeyman for all lawyers is that ChatGPT is just going to invent something in order to positively answer the question or give you the answer that you that you want. So how do you safeguard against that?

SPEAKER_03:

Aaron Powell So you know, I was actually I was stunned. I had ChatGPT, I did a presentation on these tools for a ratepayers advocate office, and I wanted to do a demonstration and show what kind of research you could do. So I had to do research and energy project. I used the deep research feature, and it was pulling cases from Public Service Commission decisions, which are incredibly obscure. They're very difficult to find. And even if you're doing research on like Lexus and Westlaw, you have to subscribe to specific libraries to get access to them. ChatGPT was pulling them up from the docket sheet and creating links to them. And then I could read and verify the results. So I was really shocked at how extensive the research was. And the other thing it did, it sort of does multi-level research. So in addition to the legal precedent that it found, it also was finding interviews with company CEOs on TV talking about the projects or talking about some of these things. It had news reports, it had links to legislative hearings. I mean, it was just this whole multi-dimensional research that was so much more robust than what you would get when you're just researching case law.

SPEAKER_01:

Aaron Powell And I'm I'm thinking like I've never had ChatGPT annotate, but I don't use the deep research very often. Right. And so the deep research does flag for you, okay, here's where I found this fact, and then you can go back and verify it. That's right. And have you built, is it a custom GPT, or is it you just feed the order in and ask it to summarize and annotate?

SPEAKER_03:

Aaron Powell I should probably come up with some consistency. Right now I'm kind of like the kid in the candy store, just stuffing my mouth full of candy.

SPEAKER_01:

You don't actually have to have a process because you could just do whatever you think works the best for any individual case.

SPEAKER_03:

Aaron Powell Yeah. But for Claude, I've set up a what do they call it? Claude, they call it a project. And so I've uploaded the data requests format, um, some of the orders in cases where issuing data requests and interrogatories, and I have it set up to generate well, that we call the data request and administrative proceedings. So I have it set up to generate data requests in the appropriate format, and then also to create objections to data requests and file uh responses or motions to compel to data requests based on documents that I've uploaded from past cases. So that's how I do it with CLOT. For ChatGPT and the deep research feature, I just upload the documents on a case-by-case basis because I mean the deep research feature takes takes a little bit longer. It's 10 minutes as opposed to 10 seconds. A little bit more of a commitment there.

unknown:

Trevor Burrus, Jr.

SPEAKER_01:

I heard I heard recently that when you asked ChatGPT a question, you know how it scrolls and it answers. It actually doesn't have to do that. It could just tell you what the answer is. But they found that when they did that, people believed the answer less authoritatively than when it actually did the animation look like it was thinking.

SPEAKER_02:

Aaron Powell Well, I've never heard that, but that makes sense. And now it even sometimes says I'm thinking.

SPEAKER_01:

Show me the thought process so that I can give you some feedback on that, and then let's get to down to the answer. Aaron Powell With what you're doing, is there a is there a client privacy and a redaction concern that you have to worry about?

SPEAKER_03:

Aaron Powell I'm not so concerned with a lot of what I do. I I handle a large number of appeals cases also, and I feel very comfortable if something's been filed in the appellate docket and it's already on the websites, I feel comfortable just uploading that as is. There really isn't any confidential client information. Once in a while there'd be an address that I might redact, but I haven't had a situation where, you know, for example, if I were representing Social Security disability litigants and they've got their social security number on everything, I might have some concerns there. But I don't have anything that broaches on that level of that requiring that kind of level of protection.

SPEAKER_01:

Yeah, we've built something similar in-house with our long-term disability practice. And we do, we have a process for going through and blacking out names and other identifying health information before we upload it. But then we've got a custom GPT with about 20 of our PDFs that takes it through our entire playbook of what does this disease process look like? What kind of what you know, review the uh policy and tell us what evidence is needed. Are there any limitations? And it spits back the kind of thing that we used to outsource to a paralegal to to run through and give you the summary of the case, condensing these often 1500 to 2,000 page um claim files into a two, three-page report. And so the concern with all lawyers is AI is gonna come and take all of our jobs. And so it's AI, it sounds like for you, has freed up a bunch of your time. And so with all of this additional time, how are you scheduling your day and what other things are you able to work on that you otherwise would have spent reviewing the eight-page order or the or uh or paying somebody to review it?

SPEAKER_03:

So I don't want to scare people. It definitely has I don't want to say cut into my business. It's made me more efficient in my business. So I don't want to be that lawyer who's just generating hours. But because I don't have a volume practice, it actually has limited some of the amount of work that I do. And even though I try to negotiate alternative fee agreements, sometimes it's just hourly. So one of the things I've been thinking about is trying to incorporate or create a new practice area that's more volume based, because I think that's where AI gives you the advantage, is it allows you to take more of those same cases that you were that you didn't have time to handle before. So that would be my solution to something like that, just because I don't know if I can really come up with the volume of cases that I would need to to outweigh or offset the time savings, which is a good thing. I mean, it's a good thing for clients. It's not something that upsets me. And I'd be I think it's a problem for me to solve as the business owner. It's not something for me to blame AI or say I'm not gonna use it. I'm gonna do my clunky billable hour times. It's an opportunity to think of new ways to practice. Trevor Burrus, Jr.

SPEAKER_01:

Well, I think there's a presentation either later today or yesterday on value-based billing and switching from the hourly. And that's a for injury lawyers and we're kind of on the bleeding edge of most of the technology stuff, because it it is the one that's easiest to run as a business, right? We think about how do we lower the cost to acquire a client, we lower the cost to run a case, AI feeds into all of that. Um and it's contingency fee-based, and it doesn't matter if it takes you eight hours or eight hundred at one.

SPEAKER_03:

That's the original alternative fee structure.

SPEAKER_01:

That's right. And so are you using it as well to create marketing and acquire more clients?

SPEAKER_03:

Yeah, I mean, like, for example, just this morning there was a notice of a rule change that was of interest to my client. So I just used AI to I kind of I read, I glanced through what the rule change was. I talked to ChatGPT and said, I want my client update to say this, this, and this, generate an update, and it did that. So and because I gave it specific instructions and it kind of glanced at what needed to be done, I'm confident that the results are the output is accurate. So yeah, so I'll send that out and see what kind of responses that generates. So that's just something I can do on the fly again that I would have never done before had it not been for AI, because I would have had to sit down and take a couple of hours out of the conference to do a marketing thing, and that might not have been worth that.

SPEAKER_01:

What is the future of your practice look like? Do you ever envision a place where you have associates or more staff or are you good right now with the work that you're doing and it's interesting and allows you to travel and speak at conferences? What does that look like for you?

SPEAKER_03:

Aaron Powell So, like I said, what I'm looking to do, and what my priority is is I've been focusing on ways that I can start an offshoot of my firm that's an AI forward practice. And whether it focuses on appeals or something like estate planning, I don't necessarily want to learn a new practice area, but I can work with somebody who's experienced in that practice area, or I can use AI to help bridge the skills gap. So I don't know need to know as much about estate planning. Well, appeals I know about, so that's something that I already know. But because I feel like if I'm going to learn something new, I mean I like to learn new things, but I'd want it rather be something non-law. I've been learning law for 30 years now. But so I see myself as having some kind of offshoot that could kind of function largely on its own and maybe turn into a saleable asset over the next couple of years, because that's one thing I haven't done with my practice. I've done all the other things. I've run it efficiently, I've become a leader in my field, I've set precedent. I mean, all the things that I set out to do, all those goals that I thought about, I've done. But the one thing I haven't done is build something scalable and sell it.

SPEAKER_01:

Yeah. What would you do? And maybe that's the answer, but what would you do differently if you were coming out of law school and setting up a law firm in 2025?

SPEAKER_03:

Well, I would definitely leverage AI tools from the get-go, and I would be very technology focused. And I guess I would look at it at the firm that I was starting. I would either try to drill down in a niche that was going to be very um that was going to be growth or a growth-oriented niche that I could build and build beyond like a single person. I think I would focus a little more on scaling. Otherwise, I would start something that's consumer-oriented and consumer-facing that could serve large volumes of people and maybe capitalize on business that people can't afford, where they currently can't afford to hire attorneys and focus on that client-based sort of um that are currently underserved.

SPEAKER_01:

Well, I look down the road and I see it frankly, I think if we put all this computing power that we're putting into AI videos, right? You can upload a script and some video of you and it creates a good enough version of you talking on video. If we put all that computing power into making cars not hit each other, I think we could probably figure out making cars not hit each other, right? Those incentives don't seem to align. And so my my fear is that that practice area up and vanishes right around the time my kids go to college. Uh I don't know whether that'll ever come true or not. Um, but I see as I drive home every day, I see people just staring at their phones. The dangerous drivers are still out there, and maybe I'm not actually at risk. But just like you, I don't really want to go back and learn any new practice area. I don't know what I would do if I had to go do something else.

SPEAKER_03:

Yeah, it's an interesting situation. I I like reading Ayn Rand, and there's she's got one short story which is about a Hollywood actress who became very famous, and she wanted to see if she could make it if she had to start from scratch, and so she goes and she becomes an unknown, and then she's like kind of challenging herself and going through like being an extra on the different scenes and she doesn't get anywhere, and she's sort of like, I wish I hadn't gone back and turned back the wheel like that. I think it would be different for you. I mean, you're always learning, and so you're always doing new things and figure out how to get somebody else to do the hard work. I think I I have kind of I I would put my money on you that you would figure out the way to figure it out. Other attorneys not so the other attorneys wouldn't be thinking about the business going away. They'd be thinking about how they can crack down and, you know, eliminate competition or keep rules in place to entrench their monopoly. So you just have a different mindset.

SPEAKER_01:

Trevor Burrus, Jr. This and this landscape is just so different than it was. When I graduate I graduated in 2008, uh events like this didn't exist where all the vendors were in one place at one time, you know. At best, you had a two-day-long state bar event sponsored by a uh court reporting company, you know, and the medical visual medical illustration company. That's another business. I don't know. I don't think you're very far from being able to upload an operative report and get a pretty good medical illustration. But coming to uh coming to a conference like this where you can speak with many, many, many people from different practice areas, different parts of the country, different businesses, I think is a real unlock for young lawyers, especially who are trying to figure out what do I do with this law degree that I have now? How many of these do you get out to a year? Are you traveling and speaking a lot?

SPEAKER_03:

I usually go, I mean, this is the first kaleidoscope conference. I usually try to get to the legal tech show if my schedule permits it. So I would say I get to three or four of them, you know, again, depending on my schedule. It is good to go out and see like what's new, what's happening in the space, what who are the new vendors are and meet people who are using the technology. So it's always good to kind of keep a pulse on what's happening so I don't get stale.

unknown:

Yeah, 100%.

SPEAKER_01:

Well, listen, I very much appreciate your time. I'll let you get back to the next session. Thank you for coming on the show. It's good to spend some time with you.

SPEAKER_03:

Yeah, thank you for having me. This is uh this has been great. I look forward to listening to the other podcasts that have come out of this.

Head Shot

Brian Glass

Host